Periodical spikes when using MCC 172 and so on

[skycraper3]

Hello,

I'm using MCC 172s but having problem now, spike noise.

 

when 4 channels connected:

blue : master board, ch0
red : master board, ch1
green : slave board, ch0
purple : slave board, ch1 (IEPE microphone connected, others are non IEPE MICs)

 

These spikes occur pretty periodically, about every 1 minute when recording signal. (regardless of IEPE status)

Can I get any help about this issue? I'm using 2x MCC172 (4Ch input), some channels spike bigger and some others not.

It seems OK when connects IEPE powered microphone with preamp but when nothing connected or using non-IEPE microphone, these noise can be noticed easily.

I'm using 5V, 3A power supply. Any helps or advices would be much appreciated.

 

Additional question, can I use mcc172_reset() in library when making DAQ boards literally 'reset' likes when RPi board freshly turns on and nothing initialized. I just want to release GPIOs that DAQ HATs holding when initialize completed (or records)

 

Thanks, Lee.

Edited November 16, 2022 by skycraper3

[Fausto]

Hello @skycraper3.

Which Raspberry Pi module and installed OS are you using?

Were the two MCC 172 DAQ HAT boards working correctly before?  If so, what changed?

Do you have other HAT boards on your Raspberry Pi module?

Do you have other equipment near by omitting EMFs?

If you enable a channel for acquisition, then it should either have a sensor connected to it or else terminated, but not left open.

Do you disable the IEPE option in the function call 'mcc172_iepe_config_write', for the non-IEPE sensors?

https://mccdaq.github.io/daqhats/c.html#c.mcc172_iepe_config_write

 

Regards,

Fausto

[skycraper3]

Thanks for reply, Fausto.

Ok, I change my question,

I connected 4x PCB IEPE Microphone in 2x MCC 172, starts recording with IEPE enabled, i got this, (its OK, of course)

 

But when IEPE disabled with same setting, I got this one,

You can see red and green noise, that is same noise as I posted yesterday.

Is it normal? If so, what is reason of that situation?

In this question, I'm on RaspberryPi 4B (5.15.76-v7l), no other HAT boards attached.

[Fausto]

Hello @skycraper3.

Thank you for the reply.  That behavior is not normal.  Which MCC 172 board and channel is represented by the red and green data?  If you swap those two cable/sensor combinations to a different channel, does the noise follow?  If the noise follows the cable/sensor combo, then the next step is to determine if the issue is with the cable, sensor or both.  Connect a 3k ohm resistor to the input channel to simulate an IEPE sensor that has a bias of 12 volts (4mA IEPE enabled).  If the noise disappears, move the resistive load to the end of the cable to test the entire cable.

If the noise remained with the MCC 172 board and channel, then the issue may be with that MCC 172 board.  Does the noise appear with both MCC 172 boards?

Regards,

Fausto

[skycraper3]

Hello, Fausto.

I'm still investigating but ye, seems board has been damaged somehow.

[Fausto]

Hello @skycraper3.

From your previous post, the red data represents ch1 on the master board, while the green data represents ch0 on the slave board.  Is this correct?  Did you try swapping cables and IEPE sensors between the channels and boards?

[skycraper3]

Hello, sorry for late reply,

I received new 2x MCC172 board today (nw, planned for another raspberrypi), so just tested.

Below plot from IEPE disabled and no microphones connected condition, 1 minute record at 2048hz frequency.
That noise still occurs every one minute, I'm guessing caused by raspberrypi that damaged or emitting something, is it possible?
With proper IEPE sensors, it works nicely.

 

On 12/1/2022 at 6:56 AM, Fausto said:

Hello @skycraper3.

From your previous post, the red data represents ch1 on the master board, while the green data represents ch0 on the slave board.  Is this correct?  Did you try swapping cables and IEPE sensors between the channels and boards?

And yes, its correct. Tried swap cables and change boards.

[Fausto]

Hello @skycraper3.

In your recent test with the new MCC 172 boards, were the input channels left open or did you connect a 50 Ohm resistor to terminate each input channel?  Can you test the MCC 172 boards with another Raspberry Pi module?  Also, do you have any equipment in close proximity to your Raspberry Pi, which is emitting EMF's?

[skycraper3]

Hello, Fausto,

- Input channels were all opened.
- Unfortunately, no spare raspberry pi atm.
- no EMF emitting devices.

There is small but audible high frequency noise when DAQ(s) attached, so guessing problem in raspberry pi power supply.
Also I've tried in my home and it made AC power noise (60hz harmonic).. 

[Fausto]

An enabled input channel should not be left open during an acquisition.  Please terminate each input channel and retest.  Do the results differ? 

Are you using an approved external power supply for the Raspberry Pi or a generic power supply?

 

[skycraper3]

OK, I tested with Raspberry Pi 4B, that noise occurs in every minute. (New Raspi 4B and 2x new 172 HAT boards)

And I tested with my old Raspberry Pi 3B, noise was gone. (Old Raspi 3B and same boards with 4B)

Is it already known issue? Can i get some help?

Edited December 15, 2022 by skycraper3

[Fausto]

Not a known issue.  Please test the two MCC 172 HAT boards with your existing RPi 4B and another approved power supply adapter.  If that does not help, then test using a different RPi 4B.  Let's rule out the power supply adapter and RPi.

[skycraper3]

I tested with 2x RPi 4B boards and 2x new power adapter, input 220V 60Hz 0.5A, output 5V 3A adapters

[heyihao]

I found the same problem when IEPE enable. The current channel is connected with a 50 ohm short circuit. 

Two MCC172 boards are installed in the current equipment. When electrical noise occurs, the other three channels also appear at the same time.

I tried to use the Raspberry Pi 3B and 4B, and replaced different power adapters and different MCC 172 boards. The problem still exists.

However, the data is normal in the IEPE disable.

Edited December 16, 2022 by heyihao

[skycraper3]

Main goal of our project is deploying system that includes 100+ raspberry pi DAQ (maybe MCC 172), this noise is still problem in verification.

Is there any news in this problem? @Fausto @JRys

[heyihao]

We tried to use higher precision DAQs in parallel to MCC 172 port for synchronous signal acquisition, and found that both DAQs would have electrical pulses.

Edited December 21, 2022 by heyihao

[skycraper3]

@heyihao This problem is currently investigating by MCC staff, also I'm still waiting response,

In short: With custom IEPE circuit, noise occurred both RPi3 and 4, but sensors work fine on NI DAQs (like USB-4432...)

Edited January 9, 2023 by skycraper3

[heyihao]

@skycraper3 Yes, we replaced the sensor and cable for testing, and found that this problem still occurs.